Using machine learning to interpret and articulate both our physiology and universal laws, Jesse Woolston puts his interdisciplinary talents to work to create emotional, multi-sensory art experiences.
Latent Colour uses machine learning, simulation, and architectural visualization to recontextualize color, art history, and natural laws of flow through physics simulation. By generating artificial intelligent depictions of art using a wide dynamic of historical data, the work informs a unique history of color, composition, and form through the lens of a machine. The use of physics simulation with the AI depictions creates a unique relationship between nature and machine, recontextualizing how the viewer sees contemporary society, natural phenomena, and technology together.
The art dataset, consisting of 30,000 public domain depictions of historic works trained with StyleGAN2, were carefully curated to show the dynamic range of abstract art. From simple forms to complex patterns, artificial intelligence optics compiled multiple perspectives of a widely practiced art form.
The computational fluid dynamics model maps the GAN generated imagery directly from pixel to fluid particle in the simulations, creating a 1:1 integration of the color and composition arrangement. The containment of the fluid model within defined space creates natural behaviors of oceans crashing on things, while also allowing for custom-built solver patterns that emulate currents under the surface of the ocean.
The culmination of the work uses architectural visualization with ray-traced graphics and real-world accurate modeling, 4K-8K resolution texturing, photogrammetry, and image and light capturing of the sky backgrounds, rendered with NVIDIA GPU graphics cards.
On The Nature of Light is an audio/visual digital tryptic, constructed with artificial intelligence, physics simulation, environmental design, and music composition. The work articulates the difference between how human physiology perceives the audible spectrum, the visible light spectrum, and the complexities of color in nature, as well as how they can be understood through audio/visual art.
Each dataset, constructed from 30,000 unique specific images of biomes–specific geographic areas notable for the species living there–uses the multiple states and time of each location in related international locations. The NVIDIA StyleGAN2-generated imagery ranges from clear skies to sunsets with dynamic light and shadows, which are used as the input source, mapped across the physics model from pixel to particle.
The physics simulation emulates patterns in nature formed from millions of years of erosion through developing a continuous, forever-evolving set of pathways. The patterns start simple, then grow more complex, emulating Icelandic glacial patterns, New Zealand rivers, and the formation of the Grand Canyon.
The music composition takes the wavelengths of color from the visible spectrum depicted in each biome and converts them to the audible spectrum. The score combines a custom musical scale and fundamental waves based on Pythagorean mathematics. The piece plays forwards and backwards infinitely, with the musical material playing back with specific dynamic intonation, creating a seamless relation between the visual and sound.
The rendering uses all photogrammetry geometry and textures, creating an accurate depiction of nature within a machine to a 1:1 representation of a real-world perspective. The central sculpture and biomes morph between each other, showing the differences in our planet’s dynamic nature.
Jesse Woolston is an artist from Auckland, New Zealand. His transmedia approach to visual and auditory arts is supported by science, machine, and the aesthetics of nature. From composing musical works for the United Nations refugee efforts, to artistically aiding the Carl Sagan Institute at Cornell University in articulating their discoveries to the greater science community, Jesse’s unique perspective on science communication with emotional considerations allow him to create unique, rigorous, disruptive, and mesmerizing art experiences.
Twitter | Instagram | Facebook
Dive into the world’s greatest achievements in AI and discover what fuels the people behind them.